Variable nfb question

[catalin gramada]

But do you really want that much drop in bass frequency response?

No, it was too much, I changed the series value from 2.2k to 4.7k for 12db nfb at 40 hz. Over 500hz is no difference between compensation/ no compensation. I have to play the speakers a little bit to see if is enough control
Thanks all.

[Helmholtz]

That is strange. I quick wired the circuit and did a quick measurement at 40 hz and need 100 mV open loop and 1 V with feedback conected to drive the amp at clipping for 60Vpk output into 8 ohm load. Almost 20 db at 40 hz, 6 db around at 250 hz...

Sorry, my bad. It should read 170Hz instead of 1.7Hz. And it's only a very rough estimation, not considering the rest of the circuit.

But do you really want that much drop in bass frequency response?

[catalin gramada]

Some general considerations:

With the values shown in your right drawing, the NFB signal will be down by 3dB at around 17Hz and by 20dB at 1.7Hz.
That doesn't look very useful to me. And I don't think your OT can handle those frequencies anyway.

Also consider that 20dB more NFB means 20dB less signal level. So if you want to damp the speaker resonance (say at 50Hz), you will loose much bass response.

That is strange. I quick wired the circuit and did a quick measurement at 40 hz and need 100 mV open loop and 1 V with feedback conected to drive the amp at clipping for 60Vpk output into 8 ohm load. Almost 20 db at 40 hz, 6 db around at 250 hz...

[Helmholtz]

Some general considerations:

With the values shown in your right drawing, the NFB signal will be down by 3dB at around 17Hz and by 20dB at 1.7Hz.
That doesn't look very useful to me. And I don't think your OT can handle those frequencies anyway.

Also consider that 20dB more NFB means 20dB less signal level. So if you want to damp the speaker resonance (say at 50Hz), you will loose much bass response.

[Chuck H]

The idea you propose is a simple solution at face value. There are always phase shifts when you create a filter. Whether this could cause more harm than good is unknown without more context. No harm in trying it.

[catalin gramada]

Yes, kind of, but want to achieve a high compensation, say 20db or more into low freq range (40-80 cps) and leave almost uncompensated the mids. Is a "presence control" but shifted all way down. I wonder if it work as simple as it is or I should be concerned by instability problems please ?

[eschertron]

Didn't you just draw a presence control?

[catalin gramada]

Hello. May I ask if is a way to get a frequency variable nfb network in a way to get considerable amount of feedback for low frequencies to improve damping factor and no compensation at all for mid and hights please ? I figured some ideea but not to sure it is useful to worth to dig on it. Any thoughts please ? Thanks.